SENSE-ORGANS IN MEDUSAE 
771 
of nerve fibers is, however, most noticeable at this point. A 
thickening of this layer seems to run along the under lip of the 
pocket. The tips of the pockets extend each one into a prominence 
on the dorsal wall of the gastric pouch at the side of the rhopalial 
canal, just as in the adult Chrysaora. 
In a specimen about 26 mm. in diameter in which the tenta- 
cles of the second set are over a centimeter long the dorsal groove 
is a saucer-shaped depression lined in its deeper parts by a dis- 
tinctly columnar epithelium. The sensory niche is also deeper, 
the rhopalial ridge longer, and the lateral pockets more nearly 
filled with the Ditted epithelium, which in turn has a much less 
embryonic appearance than in the younger specimen. 
These specimens should be compared with young Chrysaoras 
of about the same size (fig. 1) rather than with the younger ones 
(figs. 33 and 34) that I have described as in the beginning of the 
adult stage. A Chrysaora 25 mm. in diameter has its gonadia 
formed, although they are quite small, and is otherwise inter- 
mediate between the beginning of the adult condition and its 
completion. In the young Chrysaora of 25 mm. (fig. 1) the thick- 
ened epithelium does not cover the rhopalial ridge but forms a 
simple band on each side until it reaches nearly to the mouth of 
the pocket, where the first pits appear. The pitted epithelium 
lines the angle of the pocket and fills one-half or two-thirds of 
it. The pockets are smaller and the pits less numerous than in 
the Dactylometras 18-25 mm. in diameter, described above, 
and, moreover, there is no extension of the pitted epithelium 
outward along the floor of the pocket, which is lined by the same 
kind of epithelium as the roof. 
Dr. Brooks has figured a young Dactylometra which has its 
third set of tentacles just budding. This specimen is 70 mm. 
in diameter and is therefore considerably older than my 
specimens, which are probably in the first half of the Chrysaora- 
stage and my comparison of them with the condition in Chry- 
saora of about the same size is the proper one. It would appear, 
then, that in the Chrysaora-stage of Dactylometra the structures 
in the sensory niche are more advanced, while the dorsal groove 
is, if anything, less advanced than in the same stage in Chrysaora. 
